Not sure if it can do video conferences with E2E encryption, because that could be very difficult (impossible?) or bandwidth intensive.
Now I just need to convince everyone to switch ;)
I could convince more to switch if chat managers like gajim/pidgin etc. would support it out of the box.
It's a good start.
From what I read, they are only matrix clients, not vector clients.
Matrix itself is just a spec / API for sending and receiving messages. Tensor and Quaternion are two other clients that work with a given Matrix home-server. AFAIK all of these clients can utilize Matrix APIs to some degree, however I'm unsure about video conferencing as I only really use Vector. That said though, other clients' capabilities don't really depend on Vector at all, unless they're using https://vector.im as an identity server.
There are a few problems, though. One is that NAT traversal will still probably mean that it will be unreliable unless you also set up a service to carry data. The other main problem is that nobody will pay you for it. There are literally hundreds of failed startups who realised how easy it is to build these things. They couldn't find a way to make money.
It should be easy because if people are forced to pay for a service to carry data in order to make it reliable, you should be able to make money. Right? The problem is that people won't pay for reliability because they think that the software is buggy (see comments above re: "I tried Hello, but it was buggy. No audio, etc." Classic).
Skype famously solved this problem by being evil and routing it's traffic through un-NATed users. Google solves this problem by having more bandwidth than god (and hangouts still sucks half the time).
But if you ignore the "there's no way to make money at this" problem, it's a pretty fun project to write.
* Where messages are de/encrypted readable for every device/endpoint of an account not just for the one with the active session.